GENERAL
PHYSICS
I.
MOLECULAR BEAMS*
Academic and Research Staff
Prof. J. R. Zacharias
Prof. K. W. Billman
Prof. J. G. King
Prof. C. L. Searle
Dr. K. Fokkens
Dr. S. G. Kukolich
D. Babitch
F. J. O'Brien
M. A. Yaffee
Graduate Students
C. M. Bell
T. R. Brown
A.
OPTIMIZATION
R. Golub
D. E. Oates
OF MODULATION FREQUENCY IN SQUARE WAVE
PHASE MODULATION
In our continued program of investigating both theoretically and experimentally the
properties of Square-Wave Phase Modulation (SWPM) 1,'
in atomic clocks, an experi-
ment has been performed to determine the response of the system to the frequency of
modulation. The results indicate that a somewhat higher frequency is desirable than was
initially predicted.2
A second result shows some deviation from our first-order theory.
=108.9
.0f
Hz
f = 74.9 Hz
Q(VOLTS)
I
f
0.75 -
0
=
200.8 Hz
f = 52.8 Hz
0.50 -
0.25
O
-100
-80
-60
-40
-20
20
40
60
40
80
100
(DEGREES)
-0.25 -
-0.50
-
-0.75
-
-i.O
Fig. I-I.
Variation of quadrature output with RF cavity phase error
for various modulation frequencies.
This work was supported by the Joint Services Electronics Programs (U. S. Army,
U. S. Navy, and U. S. Air Force) under Contract DA 28-043-AMC-02536(E).
QPR No. 86
(I.
MOLECULAR BEAMS)
Figure I-1 illustrates a few of many quadrature signal response curves that were
taken at various modulation frequencies. To obtain the data, the clock in-phase system
was locked (I= 0) at the center of the resonance.
The quadrature output, Q,
was then
measured with a synchronous detector whose phase had been adjusted to be 90* relative
to the I-channel synchronous detector. A single curve on the graph represents this output for a given modulation frequency as a precision phase shifter varied the relative
phase between the RF cavities in the microwave system.
such curves have been obtained not only for the 4,0
also for other o- and
Tr
It should be mentioned that
3, 0 hyperfine transition but
--
cesium transitions.
aQ
A plot of the Q-loop sensitivity, S
, obtained from the slopes in Fig. I-1, is
0
shown in Fig. I-2.
It is seen that for the particular velocity distribution present in our
17.5
-
EXPERIMENT
---
THEORY
12.5 -
5.0
2.5
0
I
I
I
I
I
50
100
150
200
250
I
300
fMOD (H,)
Fig. 1-2.
Quadrature sensitivity as a function of
modulation frequency.
apparatus (most probable velocity = 1 6 0 m/sec) and the dimensions (RF cavity separation
L = 128. 7 cm, detector-to-cavity drift region L1 = 100. 7 cm) the maximum occurs at
approximately 110 cycles.
and M.
A theoretical curve,
based on the calculations of D. Babitch
Bell, shows good agreement with the data except for the interesting "structures"
at higher modulation frequencies.
This phenomenon is still under theoretical scrutiny
and is possibly associated with the RF field's making an integral number of phase steps
QPR No. 86
(I.
MOLECULAR BEAMS)
in the atom transit time between cavities.
Figure I-3 illustrates the sensitivity of the I-channel SI ulation frequency.
aI
as a function of mod-
Note that the resonance linewidth was 60 Hz with a modulation
60 F-
50
40
30
20
0
I
I
I
50
100
150
I
200
250
300
fMOD (Hz)
Fig. 1-3.
In-phase channel sensitivity as a function
of modulation frequency.
frequency of 52. 5 Hz, and increased monatonically to 74 Hz at 250 Hz.
The optimum
S I is seen to occur at approximately 90 Hz. Thus a reasonable, and convenient, operating
point which balances the two channel sensitivities will be at fmod
mod = 100 Hz.
K. W. Billman
References
1.
R. S. Badessa, V. J. Bates and C.
p. 175, January 1964.
2.
S. G.
B.
THE TWO-CAVITY
L. Searle, IEEE Trans. Vol. IM-13,
Kukolich and K. W. Billman, J.
The J = 3,
Appl. Phys. 38,
No.
1,
1826 (1967).
MASER AS A POSSIBLE FREQUENCY STANDARD
K = 2 inversion frequency of NH 3 has been measured,
hyperfine transition frequency used as a reference.
with the cesium
These measurements were made
in order to determine the frequence stability and resetability of the two-cavity maser
system.
The short-term stability over a period of 20 minutes with 20-sec averaging
time was 3. 2
QPR No. 86
10 - 11 (standard deviation).
The standard deviation over a 3-hour period
(I.
MOLECULAR BEAMS)
with 200-sec averaging time was typically 7 X 10 - 11.
on different days was 5 x 10
-11
The standard deviation for runs
for a 3-day period.
These measurements indicate that
this device could be developed into a useful frequency standard. The physical construction of this device has been described elsewhere.1, 2
The molecular resonance Q of this device (frequency/line-width) is 7 X 107, and the
signal-to-noise ratio is 104 (1-Hz bandwidth) for this line. A signal-to-noise ratio of
105 (1-Hz bandwidth) has been obtained on the N 1 5 H3 3-3 line.
-12
stability of 10
, or better.
This indicates a potential
The primary contribution to instability in the present system is the phase difference
between the RF fields in the two cavities. This phase difference is monitored by using
a square-wave phase modulation system 3 and corrected with a phase shifter. The error
in setting this phase probably accounts for the measured stability. In spite of temperature stabilization of the cavities, the temperature of the cavities and waveguide system
fluctuate slowly and introduce phase shifts that must be compensated.
The frequency of the central component of the 3-2 inversion transition was 22, 834,
184, 960. 0 ±1. 1 Hz for the 3-day series of measurements. These measurements were
made with respect to the Al time scale which locates the cesium (4, 0- 3,0) zero field
hyperfine transition at 9, 192, 631, 770. 0 Hz. The reference system for these measurements was a cesium atomic beam apparatus with 70-Hz linewidth. The reference frequency standard was operated with a C field of 82 mg.
S. G. Kukolich
References
'1. S. G. Kukolich, Phys. Rev. 138, A1322 (1965).
2. S. G. Kukolich, Phys. Rev. 156, 83 (1967).
3. S. G. Kukolich and K. W. Billman, J. Appl. Phys. 38, 1826 (1967).
C.
INVESTIGATION OF THE VORTEX STATE IN TYPE II SUPERCONDUCTORS
BY MEANS OF AN ATOMIC BEAM
The vortex state in type II superconductors has been observed by both neutron diffraction and nuclear magnetic resonance, but only with moderate resolution. Also, these
techniques cannot be extended to observation of the predicted motion of the vortices when
a current is flowing in the superconductor. The proposed experiment will be able both to
increase the resolution and to provide a technique for observing the theoretical vortex
motion.
The experiment will be performed by passing a state-selected beam of atoms near
the surface of a superconductor. By observing the velocity distribution of the flopped
atoms, it will be possible to obtain the spatial variations in the field distribution at the
QPR No. 86
(I.
MOLECULAR BEAMS)
surface, since, to atoms moving at velocity v, a spatial variation of magnetic field of
wavelength f will appear as a time-variant field of frequency f = v/l. Therefore knowledge of f and v will lead to f.
The motion of the vortices can be observed as a shift in
the velocity distribution of the flopped atoms, provided samples of sufficient purity can
be obtained so as to avoid pinning.
During the last quarter, the construction of the experimental apparatus has been
completed, and the apparatus assembled. The Helium dewar has been leak-checked for
both vacuum and heat leaks.
The heat leak is 50% less than the limit allowed in the
The gain of the electron multiplier has been measured and found to be
at 200 volts/stage with 15 stages. The C-magnet has been calibrated and the field
designing stage.
10
is within the needed range.
Wider slits have been put into the potassium oven, in order
to get a higher beam intensity. This has been observed in a measurement of 7 X 109
atoms/sec of induced low-frequency flop with 10% direct beam.
The necessary electronics equipment is now being designed. This consists of various preamplifiers, coincidence gates, integrating circuits, etc. A GR 1392-A pulsedelay generator will be used to measure the time of flight of the atoms. When the
electronics is completed, it will be possible to maximize the signal-to-noise ratio for
a selected velocity range.
T. R. Brown, J. G. King
D.
QUARK SEARCH
An experiment designed to search sea water samples for quark ions has been performed. The basic assumptions will clarify the underlying principles involved in the
experiment.
Basically the experiment consists of taking a sample of sea water (obtained from
Woods Hole, Massachusetts) and concentrating the ion and quark-ion content by evaporation of water. Such evaporation is done in a typical hot-water heater, and the concentration of the sample determined by flame photometry. For much of the experiment
described herein, a 10:1 concentration is used.
The sample is then put into vapor form by a two-step process consisting of atomization and evaporation.
The atomizer used is a pneumatic type which breaks the
sample into very small droplets. The droplets are then evaporated in a flame which is
swept of negative particles by an electric field, thereby collecting ions and quark ions.
Since the collecting surface is at a positive potential, the regular negative ions will
become neutral or form oxides.
The quark ion will maintain its net charge of -1/3,
regardless of its reaction history. The collected quark ions are then introduced into a
detection system consisting of an electron multiplier capable of detecting single particles. The time of release of the quark ion from the collecting surface is determined
QPR No. 86
(I.
MOLECULAR BEAMS)
Table I-1.
Results of analysis of 10 runs.
Actual sea water sampled
Equivalent sea water sampled
12. 3 moles
123
moles
7. 38 X 1025
Number of molecules sampled
Analysis
Average background during runs
m = 0. 52 counts per
observation period
Background to 90% confidence
3 y = ±3 \m
= 0. 3 counts
Efficiency of the detection system
0.5
Minimum number of quark ions which could
be detected per run
8
Total particles detected
<80
Limit: No. quark ions
1-<.08 x 10 - 24
No. water molecules
No. quark ions
<6 x 10 26
No. nucleons
No. quark ions
<36
liter
by controlled heating of the surface, as well as by controlling the potential restraining
the quark ions.
The experiment is based on the following assumptions:
1.
A quarked atom or water molecule has a permanent net charge of -1/3.
2.
Because of the solvation energy associated with an ion in water solution, a quark
ion will remain in solution at temperatures of 100°C.
This allows the concentration of
quark ions by evaporation.
3.
The energy required to remove a quark ion from solution is adequately supplied
by a pneumatic atomizing, flame evaporation system.
4.
Once the quark ions are free of the sea water, they may be collected by an elec-
tric field.
Since they have a fractional charge, they will not be neutralized by a posi-
tive probe, but held as ions.
5.
The neutralization of three quark ions by an integral charge is an unlikely event.
QPR No. 86
(I.
6.
MOLECULAR BEAMS)
The collected quark ions may be released by means of controlled heating and
favorable potential conditions.
7.
The quark ions can be counted with an efficiency of
100% by an electron multi-
plier and counting system.
these
Since
assumptions
are
pertinent
for a
successful
search,
each
must
be
justified.
Results
Preliminary runs designed to look for relatively large numbers of quark ions with a
sensitivity of approximately 100 particles per run showed less than 100 quark ions per
mole of regular sea water.
Later runs with a total system efficiency estimated at 50% sampled 12. 3 moles of
10:1 concentrated
sea water.
Each run consisted of an 18-msec observation period
during which counts were recorded, and from three to six 18-msec periods which served
to determine
order
background during the run.
of 1600
counts/min,
the
Although the actual background was of the
background
during the
18-msec
run time
varied
between 0 and 4 counts.
The
10 runs are analyzed on a cumulative basis and the results are given in
Table I-i.
J.
QPR No. 86
M.
Franz, J.
G. King